KR101521324B1 - Tracking area management method and apparatus for long term evolution telecommunication systems - Google Patents

Abstract

The MME may determine a TA transition probability matrix (" TA ") derived from a global table holding data of UE movement in the network by taking into account the current TA of each UE and the most recently known previous TA for all TAU events and paging events transition trait mobility trajectory by periodically updating the transition probability matrix. The MME also maintains data regarding the paging event and the number of TAUs performed by each UE and stores the paging ratio for each UE's TAU. The UE characteristic, the UE paging rate, and the network mobility property are used to minimize the overall traffic cost function for the paging event and the TAU event for the UE and the entire network in an algorithm that creates a TA list for each UE. Optionally, the TA list for each UE may be constrained to meet certain minimum performance characteristics, such as a predetermined paging success rate target and / or a predetermined delay limit target.

Description

TECHNICAL FIELD [0001] The present invention relates to a tracking area management method and apparatus for a long term evolution communication system,

Cross reference of related application

This application claims priority from U.S. Provisional Patent Application No. 60 / 150,499, filed February 6, 2009, the entire contents of which are incorporated herein by reference.

The present invention relates to tracking area management in an LTE (Long Term Evolution) communication system.

The third generation partnership project (3GPP) has developed a specification for improvement in wireless communication systems, often known as LTE (Long Term Evolution). LTE has many improvements and advances over the previous generation of wireless communication networks and systems. Among them, there is dynamic tracking area management. In particular, user equipment (UE), such as mobile phones, laptop computers, wireless personal digital assistants (PDAs), etc., is mobile by definition and can move between cells over time. Accordingly, a wireless communication network typically has a technique or protocol that maintains data regarding the location of the user equipment for that network.

The LTE specification describes a protocol that maintains data about the location of a UE on a network. In particular, LTE provides dynamic management of UE location.

In this specification, it is assumed that LTE is basically known. In LTE, the UE interfaces with the network via an evolved node B (eNB). An MME (Mobility Management Entity) at the main signaling node in the network is responsible for initiating paging and authentication of the UE. The MME also maintains the location information of the UE.

LTE introduces the concept of a tracking area (TA). The tracking area is a subset of the volume of space in the wireless network where any given UE may be located. The tracking area may include an area covered by one eNB (e.g., cell) or multiple eNBs (multiple cells).

According to the LTE specification, the MME knows the location of the UE at the TA level granularity when the UE is idle (e.g., when it is not actively communicating over the network, as in an active telephone call). Each UE maintains a Tracking Area (TA) list that may include one or more TAs where the UE may be located. Only when the UE leaves the area covered by the TA in its TA list, the UE initiates a tracking area update (TAU) operation to notify the MME of its new location. In response to the TAU, the MME typically returns the updated TA list to the UE.

In summary, the tracking area update is the communication between the UE and the MME (e.g., via the eNB) that informs the MME of the new tracking area of the UE. The MME may also send data to the UE in connection with tracking area management.

When a call is made to the UE (e.g., a voice call to the cell phone), the UE is paged by the network at the TA in the assigned TA list that the UE last knew. As a result, if the UEs in the network tend to have a large TA list, the TAU traffic level tends to be relatively low, but the paging traffic level tends to be relatively high. In particular, the greater the number of TAs in the list, the more likely the UE is in the area covered by the TA in its TA list. Thus, the UE may perform TAU less frequently. On the other hand, if the TA list is kept relatively small, there will be many TAU traffic but less paging traffic. In particular, if the UE's TA list is small, there is a relatively high probability that the UE will leave the area covered by the TA in the TA list, and therefore the TAU will need to be performed more frequently. In addition, since the number of TAs in the list is small, every time the UE is paged by the network, the number of TAs that the UE must potentially be paged before the location of the UE is verified is small.

Previous generation wireless network technologies, such as Global System for Mobile communication (GSM), used a static routing area or location area management mechanism, which caused complicated offline network design problems. In addition, the network mobility characteristics that change over time over the operating lifetime of the network, even if well engineered at the time of network design, can quickly make the network design less than optimal for a given use of the network. In addition, this static tracking area management mechanism may not be adapted to yield optimal signaling load results for each individual UE. Therefore, regardless of the change in network mobility characteristics, the performance of the static tracking area management mechanism is still lower than that of the dynamic tracking area management enabled by LTE.

According to the present invention, the MME derives from the global table which keeps data of the UE movement in the network by taking into account the current TA of each UE and the most recently known previous TA for all TAU and paging events Track the network trace mobility characteristics by periodically updating the TA transition probability matrix. The MME also maintains data regarding the paging event and the number of TAUs performed by each UE and stores the paging ratio (ratio of paging to TAU) for each UE. UE characteristics, UE paging rate, and network mobility properties are used to minimize the total traffic cost function for paging and TAU events for that UE and the entire network in an algorithm that creates a TA list for each UE. Optionally, the TA list for each UE may be constrained to meet certain minimum performance characteristics, such as a predetermined paging success rate target and / or a predetermined delay limit target.

1 is a conceptual diagram of an LTE network including a plurality of tracking areas;
Figure 2 shows a transition probability matrix M according to the principles of the present invention.
FIG. 3 shows a version P normalized for each row of a transition probability matrix M. FIG.
Figure 4 shows a further modified version Q of a transition probability matrix M;
5 is a graph illustrating how the cost function L _i typically varies as a function of the number of TA in the TA list n _i , in accordance with the principles of the present invention.
6 is a flowchart illustrating an operation according to an embodiment of the present invention.

1 shows a block diagram of an exemplary LTE network including ₁₂ eNBs (104 ₁ -104 ₁₂ ) each having coverage zones (or cells) (102 ₁ - 102 ₁₂ ) . As is conventional, there is some overlap between cells so that a user can move between cells without losing service or degrading service quality. The network also includes an MME 112 that is in communication with the eNB. Of course, there are many other components of the network system 100. However, this figure shows only the components that are critical to the discussion herein. In addition, the communication link between the MME and each eNB is omitted from the drawing in order not to obscure the description.

Anyway, each eNB 104 ₁ - 104 ₁₂ may communicate with the MME 112 to exchange network management information including information such as a tracking area list, UE location, and the like. To simplify the present description, it will be assumed that each zone 102 ₁ - 102 ₁₂ corresponding to the eNBs 104 ₁ - 104 ₁₂ is the tracking area TA. However, as described above, the present invention can be applied to a network including a plurality of eNBs 104 in a tracking area.

As noted above, in the LTE network, each UE maintains a list of TA's comprising one or more TAs for which the UE is registered. In addition, whenever the UE enters a TA that is not on its TA list, the UE executes the TAU.

According to the present invention, the MME includes a transition probability matrix such as the transition probability matrix M shown in FIG. 2, a normalized transition probability matrix such as a normalized transition probability matrix P shown in FIG. 3, And maintains an ordered transition probability matrix, such as an ordered transition probability matrix Q, in the computer memory. In particular, the transition probability matrix M includes the sum of the tracking area management event counts in the network. The tracking area management event is, for example, a paging event and a tracking area update (TAU) event. The table is updated at predetermined intervals such as every week. The value in each cell of the matrix M corresponds to the number of UEs repositioned to a new TA (current TA) represented by the corresponding column number from the TA (most recent previous TA) represented by the corresponding row number. These numbers may represent, for example, a combined sum of the UE-initiated TAU and the MME-initiated UE page. For example, according to the table, 187 UEs have moved from network cell 102 ₆ to network cell 102 ₂ , 213 UEs have moved from cell 102 ₈ to cell 102 ₃ , and 0 The UE has moved from the cell area 102 ₅ to the cell area 102 ₈ , and so on.

This matrix may be newly generated for each interval based on only the TA tracking area management event that occurs after the last update interval, or may include moving window editing of data including both new data and data from a predetermined number of previous intervals moving window compilation). The network operator can choose a way that is thought to be likely to provide data that better predicts the future movement of the UE in that particular network. It may be desirable to apply an exponential weighting factor l (where l is 0 to 1) in order to ensure that the numbers do not become unnecessarily large, especially when the moving window scheme is chosen because the event count is quite large It can be.

In general, when data implies a slow time varying network mobility characteristic, λ should be chosen to be close to 0, and when data implies a fast time varying mobility characteristic in the network, λ is 1 . Thus, assuming the use of an exponential weighting factor, the exponentially weighted value that is filled in the cell of the transition probability matrix M can be expressed as:

Where m _ij is an exponentially weighted value at column i, row j,

lambda is an exponential weighting factor,

t is the time, and u _ij is the number of UEs that have switched from TA _i to TA _j in the associated period t.

It should be noted that the number on the diagonal of matrix M is not all zero because there may be situations in which conversions may be recorded even if the UE remains in the same TA. For example, the UE may simply perform periodic TAU or similar reporting operations, regardless of whether it has moved or not. It is also noted that, in many, if not most, real-world networks, it is statistically possible that the UE may remain in the same TA between any two periods, and this fact is not correctly represented by the number in the matrix M.

The normalized transition probability matrix P includes transition probability data obtained by normalizing the data in the matrix M for each row sum on a row-by-row basis. The obtained probability data (i.e., columns) are then sorted in descending order. That is, the values in the cells of each row of the matrix M are such that the sum of the numbers in each row of the matrix P is 1 (before rounding), and the values in each cell essentially correspond to the row numbers Divided by the sum of all the digits in the row so that there is a probability of switching from the TA to the TA corresponding to the column number. Then, the columns are rearranged in descending order according to the probability value. In Figure 3, there are two numbers in each cell. The first number is the above probability value. The second number (in parentheses) is the number of the TA that the UE is switching to (ie, the row number in the matrix M from which the probability value came). For example, a cell in row 3, column 5 of the matrix P of FIG. 3 indicates that the probability that the UE will move from TA3 to TA8 is 0.11, which is the number 198 in column 8 of row 3 of the matrix M. [ Is obtained by dividing the sum of matrix 3 by the sum of row 3 (1879).

To further facilitate later calculations, another matrix, the ordered transition probability matrix Q, is defined as:

Where q _ij (t) is a value in the cell corresponding to column i, row j, for time t,

N is the number of TA in the network,

p _ii is the value in row i, column i of the normalized transition probability matrix P.

Matrix Q uses the same notation as mentioned above for matrix P.

Note that q _ij (t) = p _ii (t) when j = 1 is essentially a condition to place the diagonal value of P in the first column of Q. In essence, converting a matrix P to a matrix Q is merely to move a diagonal cell of the matrix M (this cell represents the transition from any given TA to exactly the same TA) to the leftmost column in each row , All other cells in the row are shifted to the right by one row as needed to accommodate the move. This is because, for each TA, the UEs in that TA from the left to the right are statistically based on historical data recorded in the transition probability matrix M (and even if these events are not normally recorded in the matrix M, The matrix Q that lists the TAs most likely to be found during the next time interval is generated.

In addition, let us define the re-ordering index matrix V as:

Where v _ij (t) is the value of row i, column j of matrix V at time t,

k is the number of TA in the TA list.

The matrix V maps the matrix Q back to the matrix P.

Note that for a TA of size K, the probability (hereinafter referred to as "TAU probability") that the UE in the TA performs the TAU is equal to Equation 4:

If each row in the matrix Q is expressed by Equation (5)

의 열에 대응하는 TA는 TA 목록에 있어야만 하는 TA이다(왜냐하면 이들이 UE가 발견될 K개의 가장 유망한 TA이기 때문임). 한편,

의 합은 추적 영역 업데이트 확률이다.(Where K is the number of TA in the TA list of the UE in the corresponding TA)

The TA corresponding to the column of the TA must be in the TA list (because these are the K most promising TAs the UE will find). Meanwhile,

Is the tracking area update probability.

It should also be noted that when the TA list is updated, the TA presently present in the UE must be included in the UE's TA list, regardless of the magnitude of the probability. Otherwise, the TAU will be triggered immediately. This is because the first column of matrix Q is the diagonal of matrix P.

As will be seen below, matrix Q will be used in the algorithm to derive the TA list for all UEs in a given TA that will minimize the overall network traffic for performing TAU and UE paging.

In addition to maintaining data regarding the entire paging and TAU in the network and updating the transition probability matrix accordingly as described above, the MME also tracks the number of times the paging is performed and the TAU is performed for each individual UE . The MME calculates the paging rate at each UE in all data collection time intervals t. The paging rate is as follows:

A small positive sum added to the denominator is to prevent the possibility of dividing by zero if there is no TAU during the relevant period.

Thus, g varies in proportion to the size of the TA list (i.e., the number of TAs in the TA list). (More specifically, the larger the TA list, the less the number of TAUs performed by the UE and the greater the number of paging performed by the eNB.) Alternatively, the exponent weighting factor may be set to a paging ratio g (t) .

Let's define two or more values as follows:

As can be seen, β (t) is the real-time cost of the TAU event divided by the real-time cost of the paging event. Network operators can define real-time costs as desired. A reasonable definition of the real-time cost of a TAU event or paging event is the average CPU load required to perform it. However, it may also be defined as the amount of average data transmitted or the average amount of network broadcast time consumed by this event.

The paging success rate r may be defined as the ratio of the number of times that the paging for the UE establishes contact with the UE for the total number of paging times.

The exact algorithm for generating a TA list for a UE in accordance with the principles of the present invention to minimize overall aggregate traffic for paging and TAU events will, of course, depend on the particular paging strategy used in the network. Three exemplary paging strategies that are satisfactorily designed to contact the UE with a minimum number of attempts are discussed below. However, other satisfactory strategies are also possible, and the equations described herein below may be modified as needed for any of these other strategies.

According to a first potential strategy, the eNB first pages only the last known TA of a particular UE. If unsuccessful, the eNB pages all TAs in the UE's TA list.

If still unsuccessful, the eNB retries to page all TAs in the UE's TA list up to a predetermined number of retries D _max , and the interval between retries (hereinafter referred to as the timeout period t _d ) Increase for each retry. For example, the timeout period t _d seconds is d can be set, where d is the number of retries (d = 1, 2, ... , D max) (i. E., Is a, t _d for the first retry T _d is 2 seconds for the second retry and t _d is 3 seconds for the third retry and is the maximum D _max seconds for the last retry.

Alternatively, according to the second potential paging strategy, all TAs in the UE's TA list may be paged simultaneously at the beginning and the dth retries (where d = 1, 2, ..., D _max ) The retry (within the TA list) after the timeout period of t _d is as described above with respect to the first paging strategy.

According to a third potential paging strategy, the UE is first paged at its last known TA. If unsuccessful, the UE is paged in all TAs in its TA list, there is a maximum of D _max1 retries, each retry is done d seconds after the previous retry and the maximum D _max1 times A retry is performed. If still unsuccessful, the UE may be paged in all TAs in the network, with a maximum of D _max2 retries, each retry after a timeout of t _f seconds, where f is the sequence number of retries , that is _{f = D max1 + 1, D} max1 + 2, D max1 + 3, ..., a D _max2.

For each potential TA list size for the TA, the traffic cost function for the UE in TA _i (i.e., TA corresponding to row i of matrix Q) may be defined as:

 here,

Furthermore, it is worth noting that the different definitions of N _{page , i} , for the three different strategies discussed above _, would be:

Thus, to minimize the overall network traffic for the paging event and the TAU event, the TA list size n (i. E., For each row of the matrix Q) that yields the smallest value for the traffic cost function L _{i (i} . e., Equation 14).

If desired, n _i may be constrained by any additional conditions desired. For example, with the lowest traffic cost function L _i still satisfying some predetermined minimum average paging success rate S _target , and / or the average number of paging retries D _i to be less than the predetermined number of times D _max , it may be desirable to choose the equation 15) TA list of size n _i.

Equation 10 is computed for each i value from 1 to N as a traffic cost function, where N is the total number of TAs in the network. The term N _TAU in equation (10) is the average real-time cost of a TAU event. The factor? Is a normalization factor that normalizes the TAU cost to the cost of the paging event, as previously described in connection with Equation (7). The term N _page in equation (10) is the real-time cost of the paging event on the network. N _pages are calculated differently depending on the particular paging strategy selected for the network. As mentioned above, three exemplary paging strategies are disclosed, and an algorithm for calculating N _pages for each strategy is shown in Equations (11), (12) and (13) above.

Hence, the traffic cost function L derived for each possible TA list size is calculated as the sum of the paging cost function N _page and the normalized TAU cost function, N _TAU , for a given TA list size n _i .

For clarity, the following definitions relating to equations (10) through (14) are provided:

L _i is a traffic cost function for a list of size i,

i is the TA number,

K is an exemplary paging strategy number,

d _i is the delay limit target for the paging strategy, the average number of retries,

r is the paging success rate as described above with respect to equations (11) to (13)

q _ij is the data point in row i, column j of matrix Q,

q _im is the data point at row i, column m in matrix Q,

g is the paging rate at data collection time t for a particular UE being considered as defined in Equation 6,

n _i is the TA list size (for TA _i ).

The traffic cost function L _{i, which} is a function of the number of TA's in the TA list n _i , is a normal U-shaped graph, as shown in FIG. That is, as the number of TA's in the TA list increases from one to a certain number, the traffic cost function decreases and then begins to increase as the number of TA's in the TA list increases further. Thus, in general, the best value for n _i (i.e., the number of TA stored in the TA list of the UE on a given TA) is _{L i (n i) - L} i (n i -1) is zero, the largest _i less than or equal to n . ≪ / RTI > Thus, one efficient way to obtain the value n _i from which the lowest cost function is obtained is to create an equation that combines this condition with equation (9).

Hereinafter, it is assumed that the number of retries is limited to 1 (i.e., D _max = 1) by substituting Equation 10 and Equation 11 (that is, assuming paging strategy number 1) And that there is no constraint on the average paging success rate, deriving these equations will be described.

인 최대 n_i를 구한다.For each i

The maximum n _i .

Thus, as the table moves from left to right in any row i of the matrix Q, the last row of Equation (16) yields the desired TA list size for a particular UE, i.e., n _i , TA corresponding to row j = 1 to column j = n _i for row i.

6 is a flow chart illustrating the operation of creating a TA list for a UE in a given TA for the following conditions: (1) paging strategy 1, (2) D _max is set to one retry, And (3) Minimum average paging success rate constraint. This flow chart shows the process of determining the specific TA to place in the list for a single UE as well as the number of TA in the TA list. This process will be performed for each UE in the network as shown.

At step 301, the MME starts the process. In step 303, the MME initializes the matrices M, P, Q, and V to zero. The MME also determines, for this procedure, (1) the paging rate g, (2) the paging success rate r, (3) the weighting factor β that weighs the real time cost of the TAU event compared to the paging event, and It will need a value for the number N of TA in the serving network. The values of [beta] and [Delta] N are generally fixed values because they typically change only when the carrier reconfigures the network. However, g and r vary with time and must be calculated for each interval by the MME. Each UE will have its own g. (E. G., Once a week, or once a month) during the operation of the network, (2) at a predetermined interval during the operation of the network, (E.g., when the Olympics are held in an area serviced by the network) only when the network is activated and when a special event occurs, or (3) only once when the network is activated Is used) can be performed.

Then, at step 305, it is determined whether it is time to perform the next update of the UE's TA list. The update instance can be virtually anything. Typically, each time the MME receives a TAU event from the UE, the MME will update the UE's TA list. Thus, each TAU performed by the UE will trigger such an instance. However, the MME may also request the UE's TA list in response to (1) the expiration of a period after the last TA list update for that UE, (2) any UE is updated at any time, (3) Can be updated. In any case, whatever triggering instance, if nothing happens, the system just waits for the instance to happen. When the triggering instance occurs, the flow advances to step 307, where all tracking area management event data (e.g., TAU and paging) since the last update has the paging rate g for that UE as well as the matrixes M and P Are considered for updating.

Then, at step 309, the MME finds the TA, TA _{i_current} , at which the UE is located. At step 311, the rows of each matrix Q and V are updated. Specifically _, the values of q _{i_current, j} and v _{i_current, j} for j = 1, 2, ..., N are calculated.

Since row i of the matrix Q corresponding to the selected TA is now updated, the TA list to be used for the UE in this TA can be determined. Thus, at step 313, column number j is set to 1, ensuring that the TA list will include the selected TA itself (because, according to the definition of matrix Q, the first column of matrix Q is the same TA, TA because it corresponds to _{i_current} ). Then, at step 315, j is set to j + 1.

In step 317 _, the value of q _{i _ current , j} (determined in step 311) is compared with the value generated by equation (16). q _{i _ current, j} in this case is larger than or equal to that value, which the TA to TA number n which is a function of _i L _i of in the TA list, the still decreases, and therefore corresponds to the column j in the row i_current of matrix Q TA It should be added to the list. Specifically, since the columns in row i of matrix Q are arranged in descending order according to the probability of finding UEs previously found in TA _{i_current} in TA _j, the TA corresponding to the column in row i_current of matrix Q is assigned to the left Can be simply added to the TA list in the order from left to right. Accordingly, the flow proceeds from step 317 to step 319, where the TA corresponding to q _{i_current, j} is added to the TA list for TA _{i_current} . On the other hand, if q _{i_current, j} is less than the value calculated in equation (16), the cost function L _i increases as a function of adding an additional TA to the TA list, which means that the TA list has been completed, But instead will go from step 317 back to step 305 and wait for the next update instance.

Returning to step 319, when TA corresponding to q _{i_current, j} is added to the list in step 319, the flow proceeds to step 321, where it is determined whether the last column of row i_current has been reached. If not, the flow returns to step 315-319 to determine if another TA should be added to the TA list. If so, the TA list is completed and the flow returns to step 305 to wait for the next update instance to occur.

The schemes described herein can be implemented in the MME and do not require any support from other nodes (except to receive traffic data and transmit the TA list to other nodes in the network). In addition, the algorithm itself is computationally simple and has low memory requirements, which, when combined with a reduction in the achievable signaling traffic level, implies a much larger capacity increase for the MME.

Moreover, while the present invention has been described with reference to a 3GPP LTE network, the principles described herein are based on the assumption that a mobile node is capable of communicating with any network < RTI ID = 0.0 >Lt; / RTI >

The process described above may be implemented in a computer, processor, microprocessor, digital signal processor, state machine, software, firmware, hardware, analog circuitry, digital circuitry, field programmable gate array (FPGA) Any of these, including computers or other processors executing software stored on a medium (including but not limited to a compact disc, a digital versatile disk, a RAM, a ROM, a PROM, an EPROM, an EEPROM, Including, but not limited to, a combination of < / RTI > The data to be stored in the MME or elsewhere in accordance with the present invention may be stored in any suitable computer memory, including any of the above types of computer memory.

The flow may be substantially the same for different paging strategies and / or constraints, but the equation in step 317 will need to be modified according to the particular paging strategy and / or constraint.

Having thus described several particular embodiments of the invention, various alterations, modifications and improvements will readily occur to those skilled in the art. These changes, modifications and improvements which will become apparent to those skilled in the art are not specifically described herein but are to be considered part of the description and are to be regarded as within the spirit and scope of the invention. Accordingly, the foregoing description is by way of example only and is not intended to be limiting. The present invention is limited only as defined in the following claims and their equivalents.

Claims (34)

CLAIMS What is claimed is: 1. A method for building a tracking area list for a user equipment in a network having a plurality of user equipment that can be moved between tracking areas and tracking areas of the network,
Maintaining in the computer memory data indicative of the frequency of tracking area management events of user equipments between all pairs of tracking areas of the network over a period of time, paging events and tracking area update events;
Receiving relative cost data of a paging event for a tracking area update event in the network;
Using the processor to determine a percentage of paging events for tracking area update events in the network for at least one user equipment;
Predicting a number n of tracking areas to include in the tracking area list for the at least one user equipment based on the frequency data, the relative cost data and the ratio using a processor, Predictively decreasing the sum cost of paging events and tracking area update events of the paging event; And
Using a processor to construct a list of tracking areas for the at least one user equipment with n tracking areas,
≪ / RTI > The method according to claim 1,
Wherein said predicting comprises calculating, for each of the individual user equipments, the number n of tracking areas to be included in the tracking area list separately. 3. The method of claim 2,
Wherein the network further comprises a Mobility Management Entity (MME), wherein the maintaining, the determining the rate, the estimating and creating are performed at the MME. 3. The method of claim 2,
Wherein said creating step comprises for each individual user equipment at least n-1 tracking areas having the highest frequency of tracking area management events from a tracking area in which the corresponding user equipment is currently located to another tracking area, ≪ / RTI > 5. The method of claim 4,
Wherein said creating further comprises, for each individual user equipment, placing an individual tracking area in its own tracking area list. 6. The method of claim 5,
Wherein the creating comprises:
Creating a matrix comprising N rows and N columns, where N is the number of tracking areas in the network, each row corresponding to a previous tracking area in which the user equipment was located, Corresponding to the current tracking area being located;
Inserting into each cell of the matrix a value corresponding to the number of tracking area management events of the user equipment from the tracking area of the corresponding row to the tracking area of the corresponding column; And
Aligning each row such that the row corresponding to the same tracking area as the corresponding tracking area comes first and all other rows come in descending order as a function of the values in the cells of the matrix
≪ / RTI > The method according to claim 6,
Wherein creating comprises including tracking areas corresponding to the first n columns of a row corresponding to a tracking area in which the corresponding user equipment is currently located, in each of the tracking area lists. 3. The method of claim 2,
Wherein calculating the number n comprises determining for each user equipment the number of tracking areas i that yield the lowest value for L _i using the following equation:

Where beta is the relative cost of the paging event for the tracking area update event,
i is the individual tracking area in which the corresponding user equipment is currently located,
j is the tracking area that the user equipment can transition from tracking area i,
N is the number of tracking areas of the network,
N _{TAU, i} is the traffic cost function of the tracking area update events for the corresponding user equipment, which is a function of n _i ,
N _{page, i} is a traffic cost function of paging events for an individual tracking area i. The method according to claim 6,
Calculating the number n for each of the individual user equipments comprises determining the number of tracking areas i that yield the lowest value for L _i in the following equation,

Where beta is the relative cost of the paging event for the tracking area update event,
i is the individual tracking area in which the corresponding user equipment is currently located,
j is the tracking area that the user equipment can transition from the individual tracking area i,
N is the number of tracking areas of the network,
N _{TAU, i} is the traffic cost function of the tracking area update events for the corresponding user equipment, which is a function of n _i ,
N _{page, i} is the traffic cost function of the paging events for the individual tracking area i,
For the TA list of size n _i

ego,
Where g is the ratio of the number of paging events to the number of tracking area update events for the corresponding user equipment,
q _ij is the data showing the frequency of tracking area management events from the tracking area i to the tracking area j and the tracking area j is the tracking area corresponding to the jth column of the row i of the matrix. 10. The method of claim 9,
Wherein calculating the number n for each of the individual user equipments further comprises selecting a value n with the lowest value where L _i holds a predetermined paging success rate. 10. The method of claim 9,
N _{page , i} is a function of a particular paging strategy used in the network. 10. The method of claim 9,
N _{page, i} is selected from the set comprising:

Where r is a predetermined paging success rate,
q _im is data showing the frequency of tracking area management events from the tracking area i to the tracking area m and the tracking area m is the tracking area corresponding to the mth row of the row i of the matrix. 13. The method of claim 12,
Wherein the frequency data is reset to zero at predetermined intervals. 13. The method of claim 12,
Wherein the frequency data is exponentially weighted. In the network comprising a plurality of mobile nodes and a plurality of base nodes which can be moved between a plurality of sub-areas of the network, the mobile node sub- CLAIMS What is claimed is: 1. A method for determining a collection of sub-regions of a network to perform primary paging of mobile network nodes to reduce aggregation costs of management events,
Receiving from at least one of the base nodes and the mobile nodes data indicative of the frequency of sub-region management events of mobile nodes between every pair of sub-regions of the network, the sub- Wherein the one or more fixed nodes include paging events for paging the mobile node and tracking area update events for the mobile node to inform its fixed node of its location;
Receiving relative cost data of a paging event for a sub-region update event in the network;
Using the processor to determine a ratio of paging events to tracking area update events for each mobile node;
For each mobile node, a processor is used to estimate the number n of sub-areas to include in the sub-area list for each mobile node, based on the frequency data, the relative cost data and the ratio data Wherein the sub-region list comprises a list of sub-regions to be paged when the base station is paging the mobile node, the number of the paging events and the tracking area update events in the network being predictably reduced Sikkim -; And
Using the processor, creating a sub-region list having n tracking areas for each mobile node
≪ / RTI > 16. The method of claim 15,
Wherein the creating step includes, for each mobile node, at least n-1 tracking areas having the highest frequency of sub-area management events from the respective sub-areas to other sub-areas to the sub-area list ≪ / RTI > 17. The method of claim 16,
Wherein said creating further comprises, for each mobile node, placing an individual sub-area in its own sub-area list. 18. The method of claim 17,
Wherein the creating comprises:
Creating a matrix comprising N rows and N columns, where N is the number of sub-regions in the network, each row corresponding to a previous sub-region in which the mobile node was located, Corresponding to the current sub-region in which the node is located;
Inserting into each cell of the matrix a value corresponding to the number of sub-region management events of the mobile nodes from the corresponding row sub-region to the corresponding column sub-region; And
Aligning each row such that the column corresponding to the same sub-area as the corresponding sub-area comes first and all other columns come in descending order as a function of the values in the cells of the matrix
≪ / RTI > 19. The method of claim 18,
The step of creating the sub-zone lists includes placing sub-zones in each sub-zone list corresponding to the first n columns of the row corresponding to the individual sub-zone in which the corresponding mobile node is currently located . 19. The method of claim 18,
Calculating the number n for each of the individual sub-areas comprises determining, for each of the individual sub-areas, the number of sub-areas _i that yield the lowest value for L _i in the following equation: Including,

Where beta is the relative cost of the paging event for sub-region update events in the network,
i is the individual sub-area in which the corresponding mobile node is currently located,
j is the sub-area from which the user equipment can transition from the individual sub-area i,
N is the number of sub-regions of the network,
N _{TAU, i} is a traffic cost function of sub-area update events for a corresponding mobile node that is a function of n _i ,
N _{page, i} is the traffic cost function of the paging events for the respective sub-region i,
For the TA list of size n _i

ego,
Where g is the ratio of the number of paging events to the number of sub-area update events for the corresponding mobile node,
q _ij is data showing the frequency of sub-region management events from sub-region i to sub-region j, and sub-region j is a sub-region corresponding to the jth column of row i of the matrix. There is provided a computer program for creating a tracking area list in a network having a plurality of tracking areas and a plurality of user equipments that can be moved between tracking areas of the network,
The computer program comprising:
Computer-executable instructions for maintaining data indicative of the frequency of tracking area management events of a user equipment between every pair of tracking areas of the network over a time period, the tracking area management events comprising paging events and a tracking area update Events;
Computer executable instructions for determining a percentage of paging events for tracking area update events for at least one user equipment;
To estimate the number n of tracking areas to include in the tracking area list for the at least one user equipment based on the frequency data, the relative cost of the paging event for the tracking area update event in the network, Computer executable instructions, said number predictably decreasing a sum cost of paging events and tracking area update events in the network; And
Computer-executable instructions for creating a tracking area list for the at least one user equipment having n tracking areas
Readable recording medium. 22. The method of claim 21,
The computer-executable instructions for creating the at least one user equipment of claim 1, wherein the at least one user equipment comprises at least n-1 tracking areas having the highest frequency of tracking area management events from a tracking area currently in- ≪ / RTI > computer-readable instructions for use in a computer-readable recording medium. 23. The method of claim 22,
Wherein the computer-executable instructions for creating further comprise computer-executable instructions for placing the tracking area in which the at least one user equipment is currently located in its own tracking area list. 23. The method of claim 22,
The computer-executable instructions for making the above-
Computer executable instructions for creating a matrix comprising N rows and N columns, where N is the number of tracking areas in the network, each row corresponding to a previous tracking area in which the user equipment was located, The column corresponds to the current tracking area where the user equipment is located;
Computer-executable instructions for placing a value in each cell of the matrix corresponding to a number of tracking area management events of a user equipment from a tracking area of a corresponding row to a tracking area of a corresponding column; And
Such that the rows corresponding to the same tracking area as the corresponding tracking area come first and then all other rows come in descending order as a function of the values in the cells of the matrix.
Readable recording medium. 25. The method of claim 24,
Wherein the computer executable instructions for creating comprise computer executable instructions for placing tracking areas corresponding to the first n columns of a row corresponding to an individual tracking area into the tracking area list. 22. The method of claim 21,
The computer-executable instructions for calculating the number n are adapted to determine, for the at least one user equipment, the number of tracking areas i that yield the lowest value for L _i Comprising computer executable instructions,

Where beta is the relative cost of the paging event for the tracking area update event in the network,
i is an individual tracking area in which the at least one user equipment is currently located,
j is the tracking area that the user equipment can transition from the individual tracking area i,
N is the number of tracking areas of the network,
N _{TAU, i} is the traffic cost function of the tracking area update events for the user equipment in the tracking area _{i, which} is a function of n _i ,
N _{page, i} is a traffic cost function of paging events for an individual tracking area i. 25. The method of claim 24,
The computer executable instructions for calculating the number n comprise computer executable instructions for determining a number of tracking areas i that yield the lowest value for L _i in the following equation,

Where beta is the relative cost of the paging event for the tracking area update event in the network,
i is an individual tracking area,
j is the tracking area that the user equipment can transition from the individual tracking area i,
N is the number of tracking areas of the network,
N _{TAU, i} is the traffic cost function of the tracking area update events for the user equipment in the tracking area _{i, which} is a function of n _i ,
N _{page, i} is the traffic cost function of the paging events for the individual tracking area i,
For the TA list of size n _i

ego,
Where g is the ratio of the number of paging events to the number of tracking area update events for the at least one user equipment,
q _ij is data showing the frequency of tracking area management events from the tracking area i to the tracking area j, and the tracking area j is the tracking area corresponding to the jth column of the row i of the matrix. delete delete delete delete delete delete delete

Images